论文部分内容阅读
根据实际生物神经网络具有小世界连接和神经元之间的连接强度随时间变化的特点,首先构造了一个以Hodgkin-Huxley方程为节点动力学模型的动态变权小世界生物神经网络模型,然后研究了该模型神经元的兴奋特性、权值变化特点和不同的学习系数对神经元的兴奋统计特性的影响.最有意义的结果是,在同样的网络结构、网络参数及外部刺激信号的条件下,学习系数b存在一个最优值b*,使生物神经网络的兴奋度在b=b*时达到最大.
According to the fact that the actual biological neural network has the characteristics of small-world connection and neuron’s connection intensity change with time, a dynamic small-world biological neural network model with Hodgkin-Huxley equation as node dynamic model is constructed. The model excitations of neurons, the characteristics of the weights and different learning factors on the excitability of neurons statistical characteristics.The most significant result is that under the same network structure, network parameters and external stimuli signal conditions , The learning coefficient b has an optimal value b *, so that the excitability of the biological neural network reaches the maximum when b = b *.